G protein-coupled receptor kinase 2 (GRK2) as a multifunctional signaling hub

(1)

https://doi.org/10.1007/s00018-019-03274-3 REVIEW

G protein‑coupled receptor kinase 2 (GRK2) as a multifunctional signaling hub

Petronila Penela^1,2,3 · Catalina Ribas^1,2,3 · Francisco Sánchez‑Madrid^2,3,4 · Federico Mayor Jr.^1,2,3

Received: 6 August 2019 / Revised: 6 August 2019 / Accepted: 12 August 2019 / Published online: 20 August 2019

Abstract

Accumulating evidence indicates that G protein-coupled receptor kinase 2 (GRK2) is a versatile protein that acts as a signaling hub by modulating G protein-coupled receptor (GPCR) signaling and also via phosphorylation or scaffolding interactions with an extensive number of non-GPCR cellular partners. GRK2 multifunctionality arises from its multidomain structure and from complex mechanisms of regulation of its expression levels, activity, and localization within the cell, what allows the precise spatio-temporal shaping of GRK2 targets. A better understanding of the GRK2 interactome and its modulation mechanisms is helping to identify the GRK2-interacting proteins and its substrates involved in the participation of this kinase in different cellular processes and pathophysiological contexts.

Keywords GRK2 · GPCR · Phosphorylation · Interactome · HDAC6

Introduction

G protein-coupled receptor kinases (GRKs) constitute a family of seven serine/threonine protein kinases that specifically phosphorylate agonist-activated G protein-coupled receptors (GPCRs). Receptor phosphorylation triggers the binding of cytoplasmic β-arrestin molecules, which sterically block interaction with heterotrimeric G proteins, leading to rapid desensitization of G protein-mediated signaling cascades. Moreover, β-arrestins-bound GPCRs are targeted for clathrin-mediated endocytosis, a process that serves to de- phosphorylate, re-sensitize, and eventually recycle receptors back to the plasma membrane [1–5]. Beyond this canonical

role in GPCRs desensitization, GRKs are instrumental in triggering G protein-independent GPCR signaling through β-arrestins, since the latter are scaffold proteins for many cellular partners, able to assemble complex GPCR signa- losomes. Both the ligand-induced conformational changes and the phosphorylation pattern (called phosphorylation bar- coding) imprinted by GRKs on GPCRs could be important for determining the balance between G protein activation, GPCRs desensitization, and GRK/β-arrestin-mediated signaling [6–10].

On the other hand, it is worth noting that the impact of changes in GRK expression and functionality in cells might also involve non-GPCR targets. Thus, GRKs and, in particular, the ubiquitous and essential GRK2 isoform, are emerging as signal transducers by themselves, being able to interact with (and in many cases, phosphorylate) a variety of non-GPCR proteins [2, 5, 11–14] see also below). Consistent with such complex interactomes, GRK2 has been reported to participate in many cellular and physiological processes (cardiac contractility, cell proliferation, cell cycle regulation, angiogenesis, vasodilatation, cell migration, inflammation and metabolic homeostasis, to name some relevant examples). The changes in GRK2 levels and functionality taking place in different pathological contexts were suggested to play a key role in disease progression [11, 15–20].

Therefore, a main current challenge is to elucidate the GPCR and/or non-GPCR GRK2 cellular partners that are

Cellular and Molecular Life Sciences

* Federico Mayor Jr.

[email protected]

1 Departamento de Biología Molecular, Centro de Biología Molecular “Severo Ochoa” (UAM-CSIC), Universidad Autónoma de Madrid, C/Nicolás Cabrera 1, 28049 Madrid, Spain

2 Instituto de Investigación Sanitaria La Princesa, 28006 Madrid, Spain

3 CIBER de Enfermedades Cardiovasculares, ISCIII (CIBERCV), 28029 Madrid, Spain

4 Cell-Cell Communication Laboratory, Vascular Pathophysiology Area, Centro Nacional Investigaciones Cardiovasculares (CNIC), 28029 Madrid, Spain

(2)

involved in the physiological and pathological roles of this kinase and to better understand how GRK2 interaction networks are engaged in a stimulus-, cell type-, or context- specific manner. Such GRK2 multifunctionality appears to arise from its multidomain structure and also from complex mechanisms of regulation that determine GRK2 expression levels, activity, localization within the cell, and partner/target selection.

An overview of the complexity of the GRK2 interactome

In addition to its canonical role in agonist-stimulated GPCR phosphorylation, GRK2 has been reported to establish functional or scaffolding interactions with an extensive and increasing number of cellular partners (reviewed in [2, 11, 12, 21]. Alternative GRK2 substrates and/or interactors include non-GPCR membrane receptors and proteins (such as PDGF or EGF receptor tyrosine kinases or the ENaC sodium channel), key cellular kinases (p38Mapk, AMPK, PI3K/AKT, MEK1, MST1), signal transducers and switchers (IRS1, RKIP, EPAC1, Pin1, Patched, PDEγ, eNOS, APC, Gαq, Gβγ, phosducin; RhoA, RalA), transcription factors and modulators (Smad2/3, Period1/2, IκBα, DREAM), E3 ubiquitin ligases and chaperones (Mdm2, Cul4A-DDB1-Gβ, Nedd4-2, Hsp90), ribosomal proteins, membrane components (clathrin, caveolin), mitochondrial proteins (Mito- fusin1/2) or cytoskeleton components, and regulators (HDAC6, ezrin/radixin, tubulin, GIT1, α-actinin) (see Table 1 for the references and description of these GRK2 partners).

Since the role of GRK2 in different cellular processes and physiological functions is the result of the integration of its canonical and non-canonical interactomes, it is a key to validate these interactions and clearly determine their functional impact in relevant experimental models and to better understand the stimuli and mechanisms linking GRK2 to its potential partners in specific contexts.

GRK2 structure: a versatile multidomain protein

GRKs are multidomain molecules that emerged at early stages of eukaryotic evolution [22] and that display either ubiquitous (GRK2, 3, 5, and 6), more restricted (GRK4) or cell-type specific (GRK1 and 7) expression patterns [11, 14]. The N-terminal region (αN-helix), characteristic of this protein family and important for receptor recognition, is followed by a domain displaying homology to RGS proteins (thus termed RGS homology, or RH domain), reported to associate with G protein alpha subunits. In the case of GRK2

and GRK3, the RH domain has been shown to interact with GTP-liganded Gαq family members [23, 24]. A central catalytic domain with high similarity to other AGC protein kinases, such as PKA, PKB, and PKC, is shared by all GRKs, whereas a more variable carboxyl-terminal domain is important for the localization and translocation of GRKs to the membrane by means of either post-translational modifications or sites of interaction with lipids or membrane- bound proteins [12, 25] (Fig. 1). Membrane targeting can be achieved via prenylation (in GRK1 and 7), palmitoylation (in GRK4 and 6), a positively charged lipid-binding element (in GRK5) or binding to membrane phospholipids and Gβγ subunits via a pleckstrin homology (PH) domain (in GRK2 and 3).

To date, all mammalian GRKs (except GRK3 and GRK7) have been crystallized (see [26] and references therein).

The GRK2 structure in complex with Gβγ subunits [27], followed by the crystallization of GRK2 alone [28] and in complex with Gβγ and Gαq subunits [29], provided exciting insights into GRK regulation. The structural data placed the three defined GRK2 domains (RH, kinase, and PH regions) at the vertices of a triangle and constituted an excellent example of how multiple modular domains can integrate in a single molecule to transduce and modulate signaling events. Such structural design would allow the coordinated control of GRK2 activity and membrane targeting upon GPCR stimulation via dynamic interactions of the different GRK2 domains among themselves and with different intracellular targets [12, 25, 26].

The bilobular fold of the catalytic domain of GRK2 is very similar to that of other AGC kinases and highly con- served among GRKs subfamilies. It consists of an N-terminal small kinase lobe (aa 186–272 and aa 496–513) and a C-terminal large lobe (aa 273–475), sandwiching the catalytic cleft, which is stabilized by the interposition of the αC helix coming from the small lobe. As part of the nucleotide gate, a C-terminal extension to the large lobe, the active site tether (AST, aa 475–485), passes over the cleft contributing to stabilize the nucleotide and phosphor-acceptor binding sites in the active kinase. In many AGC kinases, transitions from inactive conformations (characterized by expanded lobes and disordered nucleotide gate and αC helix) to the active conformation involve phosphorylation of several regulatory motifs at the activation segment/loop, which is located in the large kinase lobe, and at the hydrophobic motif, which is found C-terminal to the small kinase lobe and is sometimes preceded by another important structural determinant, the turn motif. Phosphorylation of these sites coordinates the closure of catalytic lobes and stabilizes the active conformation of the αC helix [30]. In contrast, GRK2 lacks phospho-acceptor sites in the equivalent regulatory motifs, and the RH domain serves as an intra-molecular scaffold that maintains the small lobe of the kinase domain

(3)

Table 1 GRK2 regulatory and effector interactome Non-GPCR GRK2’s partnersType of interactionRegulatory consequence onCellular response and molecular impactReferences GRK2Partner Ligases and chaperones Cul4A-DDB1-GβDirect interaction and ubiquitination of GRK2Decreased protein stability↓ GPCR receptor desensitization and anti-cardiac hypertrophy, anti-hypertension

[100, 188] Nedd4-2Direct phosphorylation of Nedd4-2UnknownENAC channel stabilization?[189] Mdm2Ubiquitination of GRK2Decreased protein stability↑ Classical βAR-G protein signaling ↓GPCR receptor desensitization

[94, 96] Hsp90Direct interactionMitochondrial targeting

↑ Ca(2+)-induced opening of the mitochondrial permeability transition pore

[72] Hsp90Direct interactionProtein stabilizationKinase maturation in epithelial cells[88] Receptors and membrane proteins β-ENaCDirect phosphorylation of β-ENaCIncreased protein stability↑ Salt-scavenging function ↑ Hypertension[190] EGFRDirect phosphorylation of EGFRUnknown induced effect↑ EGF receptor signaling[191, 192] EGFRDirect phosphorylation of GRK2Catalytic activation↑ Opioid DOR desensitization ↑ Dopamine D3R desensitization[60, 61] PDGFRDirect phosphorylation of GRK2Catalytic activation↑ PDGFR phosphorylation and desensitization[59] PDGFRDirect phosphorylation of PDFGRUbiquitination and decreased NHERF binding↓ PDGFR activity ↓ PDFG- induced proliferation and migration in VSMCs

[193] IGF1RCo-ippDecreased activity↓ IGF-induced AKT and ERK activation, EGR1 downmodulation in HCC cells

[194] Cytoplasmic kinases p38Direct phosphorylation of p38Inhibition↓ LPS-mediated inflammation[195] p38UnknownActivation↑ FcεRI signaling in mast cells[196] AktCo-ippInhibitioneNOS inhibition in endothelial cells ↑Portal hypertension[83] AMPKCo-ipp and direct AMPK phosphorylation?Decreased activity↑ Follicle-stimulating hormone- and AMPK-dependent gluco- neogenesis in hepatocytes

[197]

(4)

Table 1 (continued) Non-GPCR GRK2’s partnersType of interactionRegulatory consequence onCellular response and molecular impactReferences GRK2Partner PKCDirect phosphorylation of GRK2Catalytic activation and plasma membrane translocation↑ Phosphorylation and internalization of ErGPCR-2 ↓steroid hormone 20-hydroxyecdysone signaling

[56, 57] ERKDirect phosphorylation of GRK2Catalytic modulation and decreased protein stability↓ GPCR desensitization ↑ HDAC6 phosphorylation and activation

[65, 69–71] SrcDirect phosphorylation of GRK2Catalytic activation and decreased protein stability↑ GPCR desensitization[58, 95] PI3KγDirect interactionIncreased activity↑ β2AR internalization ↑ AKT and NFAT activation and cardiac hypertrophy

[198, 199] MEKDirect interactionDecreased activity

↓ Chemokine-induced ERK activ

ation[200] MST1Direct phosphorylation of MST1Increased kinase activity↑ EGF-induced centrosome separation[87] Signaling switchers GαqDirect interactionDecreased activityPLCβ downmodulation ↓cardiac hypertrophy[63, 79, 201] RhoADirect interactionScaffolding of Raf1, MEK1 and ERK2 complexes↑ EGF-triggered activation of ERK and proliferation in fibroblasts and VSMCs

[202] Epac1Direct phosphorylation of Epac1Decreased membrane targeting↓ Rap1 activation in neurons[139] RalACo-ippInhibition↓ LPA-induced PLC activation in kidney epithelial cells[203] Phosducin and phosducin-like proteinDirect phosphorylation of phos- ducinReduced binding to Gβγ↑ Gβγ-dependent signaling[204] eIF3dDirect interactionIncreased protein stability↑ AKT/PI3K activity gallblad- der cancer cell migration and proliferation

[205] Signal transducers HDAC6Direct phosphorylation of HDAC6Increased deacetylation activity↑ Tubulin dynamics ↑ Pin1-mediated proliferation ↑ Migration/ proliferation in tumoral mam- mary cells and HeLa cells

[71, 111] Pin1Direct isomerization of GRK2Decreased protein stabilityG2/M transition[67] RKIPDirect interactionDecreased kinase activity↓ GPCR desensitization ↑cardiac contractility “uphold” Raf1 signaling

[206, 207]

(5)

Table 1 (continued) Non-GPCR GRK2’s partnersType of interactionRegulatory consequence onCellular response and molecular impactReferences GRK2Partner eNOSCo-ippCatalytic inhibition by S-nitros- ylationCardioprotection in ischemia[73, 208] APC (adenomatous polyposis coli)Co-ippIncreased Auxin/APC complex formation

↓ Wnt1- and Wnt3-induced β-catenin s

tabilization in kidney epithelial cells and osteoblasts

[209] Smad1/5/7UnknownIncreased activity↑ ALK1 signaling and angiogenesis[210] Smad2/3Direct phosphorylation of Smad2/3Decreased activity↓ TGFβ1-induced growth arrest in primary hepatocytes[211] IκBαDirect phosphorylation of IκBαDegradation↑ TNFalpha-induced NF-kappaB activation in macrophages[212] PDEγDirect phosphorylation of PDEγScaffolding of Src and Grb2 complexes↑ EGF-induced activation of ERK in smooth muscle cells[213] IRS1Phosphorylation and co-ippDecreased stability or functionInsulin resistance in different cell types[158, 214, 215] GITDirect interactionScafolding of Rac/PAK/MEK complex↑ Integrin and S1PR-induced ERK ↑migration (fibroblasts, HeLa cells)

[216] Cytoskeletal proteins and regulators, protein transport, organelle maintenance β-TubulinDirect phosphorylation of tubulinUnknownUnknown[217] α-ActininCo-ippCatalytic inhibition

↓ GPCR desensitization? GRK2 localization in s

tress fibers, sarcomere?

[218, 219] Clathrin heavy chainDirect interactionCatalytic activation↑ Agonist-induced β2AR phosphorylation[76] Clathrin heavy chain (CHC)Direct interactionEndocytic adaptor for cargoGPCR internalization[78] Clathrin light chain (CLC)Direct phosphorylation of CLCEndocytic adaptor for cargoGPCR internalization[220] Ezrin/radixinDirect phosphorylation of ezrin/ radixinActivation↑ Epithelial motility in kidney epithelial cells[221, 222] α and β-SynucleinPhosphorylation of synucleinsReduced binding to phospholi- pase D2 (PLD2)Inhibition of synuclein interaction with PLD[223] Caveolin 1 and 3Direct interactionCatalytic inhibition↓ GPCR desensitization?[24] Caveolin 1Direct interactionProtein scaffoldingIsoproterenol-induced eNOS inhibition[82] Mitofusin-1 and -2Direct phosphorylation of Mito- fusin-1 and -2Increased mitochondrial fusion activity?Mitochondrial resistance to ionizing radiation[91]

(6)

and the αC helix in a competent configuration and helps to maintain the kinase domain in an inactive, open conformation [31–33]. Therefore, activation of GRK2 appears to rely on interaction of its different domains with activating interactors (agonist-occupied GPCR, Gβγ subunits). This allows readjustment of the kinase domain contacts with the N-terminal helix and the RH and PH domains, ultimately leading to allosteric rearrangement of the AST and kinase domain closure (Fig. 1) [12, 25, 26].

Table 1 (continued) Non-GPCR GRK2’s partnersType of interactionRegulatory consequence onCellular response and molecular impactReferences GRK2Partner Transcription factors DreamDirect phosphorylation of DreamBlockade of Kv4.2 membrane trafficking↓ Peak current density of Kv4.2 channel in kidney epithelial cells

[224] Period1/2Phosphorylation of PERIOD1/2Decreased activity↓ mPeriod1 transcription ↓PERIOD1/2 nuclear trafficking higher amplitude of PER1 protein rhythms

[117] A list of selected non-GPCR substrates and interacting proteins of GRK2 is shown. See main text for details Fig. 1 Multi-domain structural organization of GRK2. The three

modular domains of GRK2 are sketched out: the N-terminal RGS (regulator of G protein signaling) homology (RH) domain (in blue), the bilobular central kinase domain (gray), and the C-terminal pleckstrin homology (PH) domain (green). The RH domain contacts both the kinase and PH domains and the RH domain–PH domain interface allows a possible route for allosteric communication. Binding of diverse molecules (proteins, phospholipids) to PH and RH domains and post-translational modifications of these domains (Y86, Y92, S670, S685, phosphorylation sites denoted in red) may affect the interface between RH and kinase lobes, and thus GRK2 catalytic activity. Residues indicated in the terminal αN-helix of GRK2 are packed against the AST loop (M17, L14, Y13, D10) and are critical for phosphorylation of GPCRs and cytosolic substrates, highlighting their role in the spatial arrangement of AST and of the kinase domain extension known as C-tail, which is required for kinase domain closure and activity. Other residues of the αN-helix are positioned away from the catalytic hinge and participate in GPCR recruitment, forming a hydrophobic patch for docking (L4, V7, L8, V11, S12) and also conveying allosteric activation to the kinase domain (D3, L4). Some of these exposed residues (D3) are involved in differential interactions with GPCR allowing binding selectivity. Phosphorylation of key αN-helix residues (Y13) or nearby residues (S29) may impact GRK2 catalytic activity and GPCR docking. See main text for details

(7)

Several studies support the notion that multiple GPCR and GRK regions are involved in their interaction [26, 34–36]. The αN-helix appears to stabilize the kinase domain closure via a process that may be regulated by GPCR binding [37–39], and truncation or point mutations of the N terminus of GRKs leads to nearly complete loss of receptor phosphorylation [40–43].

However, it is still not clear to which extent this region is directly involved in receptor binding or in facilitating kinase activity and receptor phosphorylation [26].

Besides the αΝ-helix domain, a proline-rich motif located just before the nucleotide gate has been identified in GRK2 as relevant for receptor binding [44, 45], and several studies suggest an important role for the RH domain of GRKs in receptor interaction [46–50].

In this regard, the recent structure of GRK5 with the β2AR [51] suggests that GRKs display high structural plas- ticity, showing a dynamic mechanism of complex formation that involves large conformational changes in the GRK5 RH/catalytic domain interface upon receptor binding. These changes facilitate contacts between intracellular loops 2 and 3 and the C terminus of the β2AR with the GRK5 RH bundle subdomain, membrane-binding surface, and kinase catalytic cleft, respectively [51]. Thus, the RH domain would serve as a docking site for GPCRs and help to control kinase activation via transient contacts of the RH bundle and kinase subdomains [26]. Despite all this accumulating knowledge, many questions remain regarding the determinants of GRK preferential targeting toward specific GPCRs, the mechanisms triggering GRK2 kinase activity toward non-GPCR targets, and the control of the scaffolding roles of GRK2 with different signaling partners. As discussed below, the ability of GRK2 domains to interact with activating factors might be modulated by association with additional partners and by protein modifications in response to other signaling networks, thus adapting the functionality of GRK2 to given cellular contexts.

GRK2 regulation: a precise spatio‑temporal control toward the right targets

Consistent with the complexity of its interactome, GRK2 expression levels, activity, and subcellular location are tightly regulated (Fig. 2). The relative abundance, localization, and activation status of GRK2 in a given cellular context, would directly determine GPCR signaling and desensitization and its interaction with additional partners.

Post‑transcriptional GRK2 modifications shape kinase activity and target selectivity

GRK2 is phosphorylated by other kinases at different residues (Fig. 1), underlying complex cross-talk mechanisms between signaling pathways. GRK2 is modulated by second

messenger-directed kinases downstream of many GPCRs.

PKA phosphorylates Ser685 at the extreme C terminus of GRK2, enhancing its ability to bind to Gβγ and the activated GPCR, which facilitates homologous desensitization of Gs- coupled GPCRs [52] and also of non-Gs-coupled GPCRs in a heterologous manner via GRK2 membrane targeting. Thus, selective phosphorylation of GRK2 by PKA in response to β-agonists augmented GRK2-Gαq association in response to acetylcholine stimulation of Gq-coupled M3 receptors, leading to inhibition of the Gq effector, PLCβ1 [53]. In turn, PKA phosphorylation of GRK2 is limited by the activity of PDE4 (phosphodiesterase-4), which settles a threshold for GRK2 modification by PKA in several contexts, avoiding inappropriate GRK2 membrane recruitment and receptor desensitization in non-stimulated cells [54, 55]. Phospho- rylation by PKC on Ser29 led to an enhanced GRK2 catalytic activity toward GPCRs but not soluble peptides [56].

In the light of structural data pointing that the integrity of the αN-helix of GRK2 (aa 1–30) with preservation of particular acidic residues is key for receptor interaction and phosphorylation [38, 40, 43], it is likely that phosphorylation of Ser29 improves kinase docking to GPCR by shaping prop- erly the intrinsically disordered αN region (Fig. 1). Alter- natively, Ser29 modification could exert allosteric effects in GRK2 activation guided by receptor interaction. In addition, phosphorylation of GRK2 by PKC on serine 680 has been reported to translocate the kinase to the cell membrane to desensitize the steroid hormone ecdysone-responsive GPCR (ErGPCR-2) [57].

In addition to PKA and PKC, GPCR stimulation can trigger other kinase cascades to modify GRK2 in a feedback loop, altering its membrane localization and/or catalytic activity. c-Src phosphorylates GRK2 on tyrosine residues located in the αN-helix (Tyr13) and within the RH region (Tyr-86 and Try-92) [13, 58]. Remarkably, the catalytic activity of tyrosine-phosphorylated GRK2 is increased toward both soluble and membrane-bound substrates, suggesting an allosteric effect on the kinase domain. Interest- ingly, the active site tether (AST) that contributes to stabilization of the nucleotide and phospho-acceptor binding sites in the active kinase makes contacts with the αN-helix.

It has been proposed that the side chain of Tyr-13 is packed against the AST residue Val-477, forming a hydrophobic patch required for both receptor docking and allosteric activation (Fig. 1). The Y13A mutant shows reduced phosphorylation of peptides and receptors, which is consistent with a defective stabilization of the active state of GRK2 [40].

It is possible that the gain of a negative charge in Y13 in close proximity to acidic residues (D10, E476) involved in the interaction with the AST loop might alter receptor docking specificity and substrate preferences. Residues Y86 and Y92 lie proximal to a hydrophobic interface formed between the RH domain and the kinase large lobe (Fig. 1),

(8)

and their phosphorylation could influence kinase domain closure, although this remains to be proved. Interestingly, activation of GRK2 by means of tyrosine phosphorylation can be achieved independently of GPCR stimulation and β-arrestin-mediated recruitment of c-Src [13]. PDGFRβ itself phosphorylates and activates GRK2, which then serine-phosphorylates and desensitizes PDGFRβ in a negative feedback loop [59]. EGFR also directly phosphorylates GRK2 on the reported tyrosine residues, allowing membrane recruitment and enhanced GRK2-mediated desensitization of opioid receptors [60] or dopamine D3 receptors [61] in a EGF-dependent manner. Even membrane receptors devoid of catalytic activity can also engage cytosolic tyrosine kinases to activate GRK2. TCR-activated c-Src leads to

inducible tyrosine phosphorylation of GRK2 and stimulation of GRK2-dependent CXCR4-Ser-339 phosphorylation and TCR-CXCR4 complex formation. Such GRK2-mediated complex drives the PI3Kγ-dependent recruitment of PREX1, which is required for full cytokine secretion during T cell activation [62].

In addition to increasing GRK2 activity, tyrosine phosphorylation within the RH domain also enhances the interactions of GRK2 with other proteins such as Gαq upon mus- carinic receptor M1 stimulation [63] or the scaffold protein GIT1 in response to integrin receptor activation ([64], see also below). Thus, tyrosine phosphorylation of GRK2 is a key event in receptor transactivation that endows this kinase with novel catalytic and non-catalytic effector capabilities.

Fig. 2 Multiple regulatory mechanisms allow precise spatio-temporal control of GRK2 levels and functionality. Multiple layers of modulation converge to control GRK2 dosage, localization, phosphorylation activity, and scaffolding functions toward both GPCR and non-GPCR partners in response to extracellular signals. Different signals act through transcription factors to modulate transcription of the GRK2 gene (ADRBK1). GRK2 transcript levels are finely tuned in diverse settings by miRNAs or via modulation of transcript translation by different mechanisms. Interaction with chaperones (Hsp90) guides GRK2 folding and maturation. A diverse set of cellular inhibitors and activators controls the catalytic activity of GRK2 and its context-specific repertoire of substrates or interactors. Catalytic activity can be influenced by protein–protein interactions in a positive (Gβγ, GPCR)

or negative (caveolin, calmodulin, α-actinin) manner. Similarly, phosphorylation by distinct kinases (PKA, PKC, ERK, c-Src, RTKs) or S-nitrosylation also modifies kinase activity and substrate selection.

Besides regulating GRK2 activation status, these factors contribute to compartmentalization of GRK2 activity by docking the protein in specific subcellular locations, in close proximity with particular substrates or partners. Finally, GRK2 degradation irreversibly modulates GRK2 activity and levels via ubiquitination by several ligases and proteolytic clearance by the proteasome or calpain proteases. Protein decay is frequently turned on by the same modifications triggering changes in GRK2 catalytic activity, as many phosphorylation sites on GRK2 behave also as enabling signals or phoshodegrons for ligases and proteases. See text for details

(9)

Different proline-directed kinases, which play critical roles in cell cycle, stress responses, and survival or metabolic control, converge in phosphorylating GRK2 on Ser- 670. Mitogen-activated protein kinases ERK1/2 and p38 have been reported to phosphorylate this residue in response to GPCR [65] or TLR4 [66] activation, and cyclin-dependent kinase, CDK2–cyclin A, phosphorylates GRK2 during cell cycle progression [67]. In a setting of ischemia, PI3K/AKT pathway activation promotes S670 phosphorylation through an unidentified kinase [68]. The S670 residue lies within the Gβγ-binding domain of GRK2 (Fig. 1), and its phosphorylation strongly impairs the interaction of GRK2 with Gβγ subunits, thereby inhibiting kinase translocation to the plasma membrane and kinase activity toward membrane-located substrates [69, 70]. Phosphorylation of GRK2 on S670 might alter potential paths of GRK2 allosteric activation. The RH domain has been proposed to act as an allosteric transducer domain, since the interaction of the α10 helix (aa 511–515) with the small kinase lobe may play regulatory roles similar to the SH2-catalytic domain linker in c-Src [27]. Close to this RH–kinase small lobe interface, the terminal subdomain of the RH region interacts with a cluster of residues in the PH domain that binds directly to Gβγ. This suggests that S670 phosphorylation may transmit structural alterations affecting the conformation of the catalytic domain and thus activity. Interestingly, this modification at the C terminus promotes a switch in GRK2 substrate specificity, fostering the acquisition of a distinctive competence to phosphorylate HDAC6 [71], also see below). In addition, GRK2 phosphorylation at S670 by MAPK disrupts GRK2 interaction with GIT1 [64], whereas it is required for kinase localization to the mitochondrial outer membrane via enhanced interaction with the chaperone Hsp90 [72]. In summary, S670 phosphorylation appears to play a central role in the modulation of GRK2 functional features such as catalytic activity toward GPCRs and non-GPCR substrates, subcellular localization, and the ability to interact with protein partners (see Fig. 2).

Regarding other types of post-translational modifications, GRK2 can undergo S-nitrosylation of Cys residue 340 within its catalytic domain, leading to the inhibition of kinase activity toward GPCRs [73], as well as ubiquitination of lysine residues related to the control of GRK2 turnover (see below).

GRK2 is present in different subcellular pools

GRK2 was initially identified as a cytosolic protein able to translocate to the plasma membrane upon GPCR stimulation [4, 13]. However, a significant amount of GRK2 was reported to be associated with internal microsomal mem- branes by means of electrostatic interactions between the N-terminal region of the kinase (RH domain) and an unknown protein which is an integral component of the

microsomal membrane [74, 75]. Interestingly, the interaction of GRK2 with the unidentified anchoring protein leads to inactivation of the bound kinase, what can be reverted by stimulation of endogenous heterotrimeric G proteins [75].

On the other hand, the C terminus of GRK2 directly binds to clathrin, an interaction that may facilitate the internalization of certain GPCRs, resulting in the co-localization of GRK2 with β-adrenergic receptors in endosomes during agonist- induced receptor internalization [76–78].

GRK2 activity is inhibited upon interaction with several cellular partners such as calcium-binding proteins, α-actinin, Hsp90, and caveolin (reviewed in [79]). For instance, calmodulin interacts with GRK2 at sites located at both the N- and C-terminal domains of the kinase and directly inhibits its activity [80]. Of note, these regulatory partners may act as adapters or anchoring factors, helping to preserve different pools of inactive kinase and to guide catalytic or scaffolding activities of GRK2 toward particular substrates (Fig. 2).

In this regard, the association of the kinase with caveolin-1 or -3 inhibits GRK2-mediated phosphorylation, suggesting that caveolins may play a role in controlling basal GRK2 activity [81]. However, since caveolins serve as scaffolds for a variety of signaling molecules including GPCRs, different MAPKs, and G proteins helping to compartmentalize their signaling, the association of the GRK2 with caveolin may also facilitate the interaction of this kinase with other signaling or regulatory molecules. In this regard, caveolin and GRK2 together regulate eNOS in sinusoidal endothelial cells. β2AR stimulation triggers the association of both proteins, and β-agonist-induced Tyr-14 phosphorylation of caveolin is required for such association. Interestingly, the GRK2–caveolin complex leads to eNOS recruitment and decrease of AKT-mediated eNOS-Ser1177 phosphorylation, resulting in blockade of NO production [82]. Of note, GRK2 has been reported to interact directly with AKT in liver sinusoids, thus inhibiting AKT-mediated phosphorylation of substrates [83], but whether such regulatory effect is mediated by caveolin has not been addressed.

Mitochondrial and nuclear pools of GRK2 have also been reported. Although GRK2 does not display a canonical nuclear localization sequence as GRK5 does [84], high levels of both GRK isoforms have been detected in the nuclear fraction of neuronal cells from the striatum [85]. Interest- ingly, like GRK5 [86], GRK2 has been found in centrosomes and was suggested to play a critical role in mediating EGF- dependent separation of duplicated centrosomes, although it has no role in centrosome duplication or microtubule nuclea- tion [87].

GRK2 has also been shown to interact with heat shock protein 90 (Hsp90), a known mitochondrial chaperone [88], to localize in the mitochondria and to regulate mitochondrial biogenesis when accumulated inside this organelle [89].

GRK2 mitochondrial levels increase during inflammation

(10)

or upon LPS stimulation in macrophages, helping to facilitate biogenesis and to restore mitochondrial function [90]. It has also been proposed that GRK2 localizes to mitochondria in HEK293 cells upon ionizing radiation and that that this kinase plays a protective role in this context by regulating mitochondrial fusion through interaction and phosphorylation of Mitofusin-1 and -2 [91]. On the contrary, in the context of ischemia reperfusion in vivo and in cultured myo- cytes, GRK2 localizes to mitochondria via ERK-mediated phosphorylation of GRK2 at Ser670 and Hsp90 binding, and contributes to mitochondrial-dependent pro-death signaling [72], increases superoxide levels, and impairs fatty acid oxi- dation and ATP production [92]. A better understanding of the stimuli triggering GRK2 localization to mitochondria and the impact of GRK2 dosage on mitochondrial function in different cell types and pathophysiological contexts is needed.

Control of GRK2 protein turnover

Protein degradation has emerged as the main mechanism for triggering rapid local changes of GRK2 functionality (Fig. 2). GRK2 is a short-lived protein and the chaperone function of Hsp90 plays a critical role in GRK2 protein stability and maturation in most cell types. Impaired Hsp90 activity results in degradation of GRK2 in a proteasome- mediated manner [88]. Moreover, GRK2 is rapidly degraded upon GPCR activation in an ubiquitin and proteasome- dependent manner [93]. Ubiquitination of GRK2 at the N-terminal lysine residues K19, K20, K30, and K31 marks the protein for its proteasome-dependent degradation in basal and agonist-stimulated conditions [94]. Interestingly, both GRK2-mediated phosphorylation of ligand-occupied receptors and β-arrestin engagement of the receptor complex are early common events triggered by GPCR activation, and both were found to be necessary for agonist-induced proteolysis of GRK2. Further studies revealed that cSrc- dependent phosphorylation of GRK2 was instrumental in this process [95]. Moreover, ERK-mediated phosphorylation of GRK2 at Ser670 also can promote its degradation [65].

Although both c-Src- and ERK-mediated phosphorylation of GRK2 can target GRK2 for degradation independently, ERK preferentially phosphorylates once GRK2 has been previously phosphorylated on tyrosine residues by c-Src [65].

The main E3 ligase implicated in GRK2 turnover by the proteasome pathway is Mdm2 [94]. Further characterization of the Mdm2-dependent GRK2 degradation revealed that upon GPCR activation, Mdm2 action on GRK2 is mediated by previous GRK2 phosphorylation on Ser670 but not by tyrosine phosphorylation, although c-Src phosphorylation facilitates the subsequent phosphorylation of GRK2 by MAPK [96].

Arrestins are known to recruit c-Src, ERK, and Mdm2 to the vicinity of activated GPCRs [97, 98] and these scaffold molecules are required to facilitate sequential GRK2 phosphorylation and Mdm2-mediated degradation upon GPCR activation. However, in the absence of GPCR activity, β-arrestins do not participate in the Mdm2-mediated regulation of basal GRK2 turnover, but instead compete with GRK2 for Mdm2 and suppress basal GRK2 degradation [96]. In basal conditions, c-Src is able to mediate Mmd2-independent GRK2 degradation, although the E3 ligase implicated in this process remains to be elucidated.

Thus, β-arrestins play a coordinating role recruiting kinases and/or ubiquitin ligases to GRK2 in the basal condition or upon activation of GPCRs, regulating GRK2 turnover via different pathways. Adding more layers of complexity, it is also worth noting that Mdm2 phosphorylation by AKT in response to stimulation of growth factor receptors such as IGFR leads to E3 ligase translocation to the nucleus, resulting in enhanced stability of cytoplasmic GRK2 and higher expression levels [94].

Notably, the phosphorylation status also affects GRK2 stability in a cell cycle context, thus governing the fluctua- tions in kinase levels taking place during cell cycle progression. Phosphorylation of GRK2 at Ser670 by CDK2–cyclin A and the subsequent binding of the prolyl-isomerase Pin1, are required for transient GRK2 proteasome degradation during the G2/M transition [67], although the E3 ligase involved was not identified. As Mdm2 protein expression is also cell cycle regulated, with a prominent accumulation through G2/M, it is tempting to suggest that this ligase plays a role in downregulating GRK2 in this context, but the con- tribution of other ligases cannot be excluded.

It has been reported that Gβ subunits have a non-canonical function acting as substrate recruiters for Cul4-RING ubiquitin ligases [99, 100]. In particular, a member of the Gβ family, Gβ2, specifically targeted GRK2 for ubiquitination and degradation by the cytosol-enriched DDB1-Cul4A- ROC1 ligase. Curiously, agonist stimulation of β2AR caused delayed turnover of GRK2 and up-regulation of global kinase levels in HEK293 cells through a mechanism involv- ing PKA-mediated phosphorylation of DDB1 and disruption of Gβ2 binding to DDB1-Cul4A ligase. This mechanism of feedback regulation and GRK2 stabilization evoked by β2AR stimulation appears to be in conflict with previously published data demonstrating GPCR-(including β2AR)- induced GRK2 degradation [101]. Moreover, while these authors suggest that DDB1-Cul4A-ROC1 ligase is the physiological ligase impaired in cardiac hypertrophy, heart failure or hypertension, conditions linked to adrenergic overdrive and excessive PKA activation, others have demonstrated that defective cardiac Mdm2 led to a significant increase in GRK2, resulting in severely impaired cardiac function [102].

Therefore, further studies are needed to define the main E3

(11)

ligase responsible of keeping steady-state levels of GRK2 in defined pathological contexts. Interestingly, Cul4-DDB1 is also recruited to the non-classical GPCR Smo through Gβ to promote internalization and degradation of Smo receptor and GRK2 itself, while in the presence of Hedgehog, PKA is activated and Cul4-DDB1 is dissociated from Smo, allowing receptor and GRK2 accumulation for signal transduction [99]. Based on this finding, it has been suggested that this ubiquitin ligase could play a broader role in GPCR regulation, being recruited as active receptors by receptor-bound GRK2-Gβ complexes to trigger their internalization, in a similar way as GPCR-bound β-arrestins bring Mdm2 and other ubiquitin ligases for GRK2 and receptor ubiquitination.

On top of that, other protein degradation machineries can participate in the control of GRK2 levels. Inappropriate calpain activation has been implicated in various disease states including cerebral and heart ischemia, muscular dystrophies, and chronic inflammation [103]. Proteasome-independent GRK2 downregulation takes place in several of these settings. In arthritis and in active relapsing–remitting multiple sclerosis (MS) or secondary progressive MS [104], inflammatory cytokines trigger oxidative stress-induced GRK2 down-regulation through the calpain proteolytic pathway.

Many substrates of calpains exhibit a PEST region, rich in proline (P), glutamic acid (E), serine (S), and threonine (T). GRK2 bears a putative PEST region (aa 591–615) and purified calpain is capable to promote partial proteolysis of recombinant GRK2 in a calcium-dependent manner in vitro, suggesting that the cleavage may occur between aa 677 and 678 of GRK2 [105]. Of note, the existence of a cAMP- dependent protein kinase A (PKA) phosphorylation site (S685) in GRK2 close to the calpain cleavage site raises the possibility of a regulatory interplay between protein phosphorylation and calpain-mediated degradation of GRK2 in the context of myocardial and cerebral ischemia, which are conditions characterized by a potent activation of PKA.

GRK2 regulation at the transcriptional level

Although regulation of protein stability appears to play a major role in determining GRK2 dosage, changes in mRNA levels of GRK2 have been noted in several pathological conditions, including hypertension, heart hypertrophy, and heart failure [18, 106], in the acute stage of infectious diseases, such as pneumonia and sepsis [107], or in tumoral patholo- gies [19]. However, relatively little is known regarding the mechanisms controlling GRK2 promoter activity and transcript stability or other layers of control, such as regulation by non-coding RNAs or RNA-binding proteins (Fig. 2).

Several SNPs within the promoter of the ADRBK1 gene and in non-coding regions have been identified. The rs4930416 and rs1894111 SNPs were found to be associated with differential antihypertensive drug response in black

and white patients [108]. Several of these SNPs within the gene were found to be in linkage disequilibrium (LD) in the black population, what could indicate that some of these SNPs may impact gene regulatory sequences. Although the influence of non-coding SNPs in GRK2 expression has not been addressed, in silico analysis showed that they may yield unique transcriptional factor-binding sites, resulting in potential changes in ADRBK1 regulation [109].

Analysis of the human breast cancer TCGA database shows that GRK2 mRNA is expressed at statistically sig- nificantly higher levels in tumor cells compared to normal breast tissue [110], which is in line with increased protein levels detected in breast cancer patients [111]. Interest- ingly, the ADRBK1 gene is located at the 11q13.2 band and is identified as part of the amplicon 11q13 which undergoes amplification at high frequencies in breast tumors, showing a strong association between copy number status and gene expression level [112].

In vascular cells, phorbol esters (PMA), as well as the activation of Gαq or α1-adrenergic signaling pathways, increased GRK2 promoter activity, consistent with increased GRK2 protein in settings of physiological vasoconstric- tion and hypertrophy, whereas pro-inflammatory cytokines had the opposite effect [13, 113]. Mitogenic stimulation led to increased GRK2 mRNA expression in T cells [114]

and increased GRK2 protein levels in different cell types [111, 115]. However, the identity of transcription factors downstream of such stimuli that are responsible for GRK2 mRNA alterations is largely unknown. It has been reported that several regions of the GRK2 promoter match with the response element of EGR-1, a transcription factor stimulated by PMA, and usually engaged in responses stimulated via Gαq-coupled GPCRs. EGR-1 binds to GRK2 promoter and a novel G(-43)A polymorphism (rs182084609) was identified within one of these EGR-1 binding sites regions, leading to a greater increase in EGR-1-induced transcriptional activity in the promoter bearing the G(-43) allele vs. the A(-43) allele [116].

Emergent findings point to a potential reciprocal interplay between GRK2 and the circadian rhythms machinery, based on the effect of GRK2 on the Period 1 and 2 (PER1/2) negative clock regulators [117, 118]. GRK2 acts post-translation- ally on PER1 and PER2, physically binding to both proteins and promoting the phosphorylation of PER2, what impairs the trafficking of PER proteins into the nucleus and causes sustained Per2 expression, disrupting circadian rhythms.

Interestingly, the mRNA levels of GRK2 in different tissues (such as colon and liver) appear to oscillate in a circadian manner as shown in the CircaDB database (http://circa db.hogen eschl ab.org/query ), suggesting that GRK2 may be a clock-controlled gene.

MicroRNAs (miRNAs) are short non-coding RNAs (ncRNAs) that are emerging as important modulators of the

(12)

expression of multiple genes via degradation of mRNA or by preventing translation of target genes. Increasing evidence demonstrates that long non-coding RNAs (lncRNAs) also regulate gene and protein expression via transcriptional and post-transcriptional processes [119, 120]. However, little is known about the participation of these factors in GRK2 regulation. It has been described that miR-K3, a Kaposi’s sarcoma-associated herpesvirus (KSHV) miRNA, facilitates cell migration and invasion via activation of the CXCR2/

AKT pathway by repressing GRK2 expression [121], pointing to a role of GRK2 in suppressing KSHV-associated tumor progression. Besides migration and invasion, miR-K3 also enhances KSHV latency and angiogenesis by targeting GRK2 and inhibiting its expression [122]. The lncRNA UCA1 increases the ability of gastric cancer cells to metas- tasize by regulating the stability of GRK2 protein via the promotion of Cbl-c-mediated ubiquitination and degradation of GRK2, adding another level of complexity to the regulation of the expression of this kinase [123]. The identification of miRNAs and lncRNA that are able to control GRK2 dosage in different conditions is an interesting future research avenue.

Revisiting paradigms on GPCR‑mediated GRK2 activation

The conventional flowchart in GPCR homologous desensitization starts with agonist receptor occupancy and cou- pling of the activated receptor with heterotrimeric G proteins which dissociate into Gα subunits and Gβγ dimers for downstream signal transduction. Released Gβγ subunits and plasma membrane phospholipids co-recruit cytoplasmic inactive GRK2 into the vicinity of ligand-bound GPCR, and they together allosterically activate the kinase for receptor phosphorylation, followed by high affinity association of β-arrestins with the phosphorylated receptor [2, 4]. Therefore, this paradigm implies that: (1) GRK2 exclu- sively phosphorylates GPCR in their agonist-bound state;

(2) G protein activation and hence GRK2-Gβγ complex formation is mandatory for receptor phosphorylation; (3) β-arrestin engagement only occurs upon GRK2-mediated GPCR phosphorylation. However, recent functional and structural data have called into question these principles. For instance, GRK2 can be docked directly to some GPCRs in a Gβγ-independent manner [124]. By using altered dopamine D2R receptors, prone to preferentially stimulate G proteins D2R[Gpro] or recruit β-arrestins D2R[βarr], in combina- tion with biased ligands directing wild-type D2R toward β-arrrestin in a GRK2 dosage-dependent manner, these authors demonstrated that G protein-independent recruitment of GRK2 is key for preferential D2R-biased signaling to β-arrrestin. GRK2 bound to D2R in such conditions is

active toward the receptor and its phosphorylation promotes β-arrestin mobilization. The fact that GRK2 can be recruited to wild-type D2R receptors with a rather modest contribu- tion from Gβγ-mediated kinase translocation poses the question of how kinase activation occurs in the presence of different receptor conformations, and whether the catalytic competence of GRK2 in such GPCR-only complex is the same than in the presence of Gβγ subunits (for instance, the extent and identity of phosphorylated residues in D2R[βarr]

and D2R[Gpro] could be different).

Other studies support the notion that GRK2 docking to GPCRs and activation can be uncoupled from Gβγ-mediated recruitment. Data obtained with the catalytically inactive GRK2-K220R mutant suggested that kinase activity was dispensable for plasma membrane and GPCR recruitment since this mutant is able to interact with Gβγ subunits. Thus, reduced efficacy of β-arrestin recruitment in the presence of the GRK2-K220R mutant was ascribed to the lack of GPCR phosphorylation by a receptor-bound inactive GRK2.

However, pharmacological inhibition of GRK2 with the small molecule Cpmd101 yields contradictory results when compared to GRK2-K220R in receptor recruitment assays.

Neither dopamine D2R interaction with Cmpd101-bound GRK2 nor with K220R mutant protein is affected [124], while the interaction of GRK2 with MOR opioid receptor is fully prevented by Cmpd101 [125]. It seems that for some GPCRs, GRK2 must be in a pro-active configuration for interaction and already folded in the context of a K220R mutation, which only impedes phosphotransferase activity.

In contrast, the inactive conformation of GRK2 stabilized by Cpmd101, that impedes the closure of kinase domain [126], is competent to interact with other GPCRs, suggesting that concurrent events can contribute differently in the recruitment of active vs. non-active GRK2 (Gβγ-preassembled complexes, interacting partners, post-transcriptional modifications). This diversity in GRK2 receptor docking might have implications, for instance, in whether Cpmd101 can prevent scaffolding roles of GRK2 in the desensitization of Gαq-coupled receptors.

Overall, these results suggest that GRK2 (in either active or inactive forms) could interact with GPCRs in different conformational states, and that different paths of allosteric activation could trigger GRK2 catalytic activity.

As discussed in the previous sections, the control of GRK2 activation involves dynamic interactions of its different domains and regions with GPCR, Gβγ, and lipids and among themselves, ultimately leading to rearrangement of the AST loop and kinase domain closure, and this activation process can also be influenced by post-translational modifications in the N-terminal RH domain and the C-terminal PH domain of the kinase (Fig. 1).

In addition to AST, the first 20 amino acids of GRKs are also quite disordered and, based on recent crystallographic

(13)

studies on GRK6, it has been proposed that this region may adopt an extended α-helix conformation that, in addition to serving as an interface with GPCR, would pack against the kinase small lobe and the AST, contributing to the stabilization of active conformation, and induce allosteric activation.

Thus, activated GPCRs could stimulate GRKs by ordering the N terminus and AST regions. Furthermore, the N-terminal region of GRKs could play a role in kinase domain closure, even independently of receptors [41].

Remarkably, both AST and the αN-helix regions differ broadly in their sequences among different GRKs, suggesting that they might convey diverse ways of allosteric activation. The identification of key residues for receptor docking and/or kinase allosteric activation in the N-termini of GRKs may help to understand GPCR phosphorylation preferences by GRKs or how GRK2 might be allosterically activated by non-target GPCRs to trigger phosphorylation of additional substrates. Mutational analysis combined with BRET-based protein interaction and kinase phosphorylation assays has allowed the identification of key surface-exposed residues (L4, V7, L8, V11, S12) in the αN-terminal helix of GRK2 that comprises a direct hydrophobic GPCR docking site able to provide GPCR selectivity (Fig. 1). For instance, mutant GRK2-L4A is impaired in α2AR but not β2AR interaction, while other mutations strongly increased kinase recruitment to α2AR [38, 40]. Differential impact of αN residues in GRK2-mediated phosphorylation of GPCRs and soluble peptides also indicates that some positions facing either the small lobe (A16, M17) or the AST loop (D10, Y13) (see Fig. 1), contribute to induce active conformations with different catalytic outcomes. Receptor docking sites on the αN (such as D3 or L4) also relay allosteric activation, as receptor-induced phosphorylation of soluble peptides is impaired when theses residues are mutated. Interestingly, these meticulous mutational studies also suggest that receptor kinase docking can be uncoupled from receptor-induced kinase allosteric activation. This is exemplified by the mutant GRK2-D10A which showed compromised catalytic activity toward soluble peptides and several GPCRs, despite its efficient docking, pointing to a poor active conformation [38, 40]. Intriguingly, this mutant fulfills phosphorylation of tubulin, while other αN mutants are competent to phosphorylate receptors but not soluble peptides. These findings suggest that flexibility at the interface of αN residues and the AST loop may confer biased kinase activity. Phosphoryla- tion of GRK2 or interacting partners affecting the αN region of GRK2 may not only quantitatively regulate the catalytic activity (as phosphorylation by PKC on S29 in the context of GPCR or peptide phosphorylation) but could qualitatively enable catalytically active conformations toward specific targets. This scenario becomes more plausible as the list of non-GPCR and non-membrane substrates of GRK2 grows in number and heterogeneity (see Table 1), raising questions

about how closure of kinase domain is achieved in a receptor- and Gβγ/phospholipid-independent manner.

Examples of context‑specific GRK2 multifunctionality

Multifunctional roles of GRK2 in the modulation of pain perception and chronicity

Pain hyper-sensitivity and chronic pain are disabling conditions conditioned by genetic and environmental factors. In the acute phase of pain, inflammatory mediators released by damaged tissues and catecholamines secreted by sympa- thetic afferents and adrenal gland under stress provoke sensitization of nociceptors to noxious stimuli in a process known as hyperalgesia, which in the short term is beneficial for the organism. Hyperalgesic factors directly stimulate GPCRs in nociceptor cells such PGE2 receptor (EP2) or β2AR among others, which activate cAMP and PKA-dependent pathways that reduce hyperpolarisation-activated inward currents and augment voltage-activated Na currents, thus increasing the frequency of action potential firing [127]. Although inflammatory stimuli or adrenergic stress typically result in short- lasting hypersensitivity, they can also evoke a priming effect that triggers more pronounced and persistent sensitization to subsequent hyperalgesic challenges or to noxious stimuli, thus contributing to chronic pain.

In acute hyperalgesia, GPCR desensitization curtails cAMP and PKA activation by either conventional GRK2- dependent receptor phosphorylation (β2AR) or by undefined GRK2-independent mechanisms (EP2) [128]. Notably, even though long-term exposure to PGE2 causes desensitization of receptor-mediated cAMP production and activation of PKA, PGE2-induced pain sensitization is not downregu- lated, i.e., successive challenges result in stronger and more prolonged pain responses [129]. Interestingly, inflammation produces a decrease in GRK2 levels in nociceptors, and reduction of GRK2 levels in peripheral sensory neurons by a variety of approaches enhanced intensity and duration of epinephrine- or PGE2-induced mechanical hyperalgesia [130–134]. Conversely, delivery of extra GRK2 prevented the prolongation of PGE2-induced hyperalgesia in primed mice [135]. These data strongly support that GRK2 is key in the control of pain priming and chronicity.

Transient hyperalgesia is a PKA-dependent process, while persistent pain is PKA independent and involves PKCε/

ERK-mediated pathways that ultimately phosphorylate transducer proteins and ion channels in peripheral sensory terminals, leading to their enhanced activation [136]. PKCε is a key intracellular mediator implicated in the onset of mechanical, thermal, or inflammatory hyperalgesia, and in transition from acute to chronic inflammatory pain. In naïve